Activating mutations in genes of the RAS family are the most common dominant acting oncogenic changes in human cancers. These tumors are particularly difficult to treat, and have a very poor prognosis. Mutations in RAS family members exhibit tissue specificity, but we do not know how tissue damage, inflammation and the resultant remodeling process results in early selection of cells carrying specific mutations in individual RAS isoforms. Mutations in the HRAS gene are mainly found in squamous carcinomas (SCCs) of the lung, skin, and head and neck, which share many histological and molecular characteristics and together constitute a major human cancer burden worldwide. In contrast, KRAS is the most commonly mutated oncogene in adenocarcinomas of the lung, pancreas, and colon. The same tissue specificity is observed in mouse cancer models: carcinogen-induced squamous tumors of the skin have almost 100% Hras mutations, whereas lung adenocarcinomas induced by the same carcinogen have almost 100% Kras mutations. The identification of the factors that underlie this specificity would provide important information tht may lead to tissue-specific or mutation-specific routes to cancer prevention or treatment. We generated novel mouse models in which the specificity for particular Ras isoforms in skin and lung has been reversed or eliminated, resulting in mice that develop Kras mutant skin carcinomas, Hras mutant lung carcinomas, or are completely resistant to chemical carcinogen treatment. This project will exploit these novel mouse models as well as computational network approaches to the identification of functions of Ras genes in normal tissue and carcinomas driven by either Hras or Kras. Mouse gene expression network data will be validated in human samples by integration with human gene expression datasets from squamous carcinomas of the lung and head and neck (generated by TCGA) or in primary cutaneous SCCs, in particular with human SCCs from BRAF-inhibitor treated patients which have an elevated frequency of HRAS (Q61L) gene mutations - exactly the same point mutation that is found in Hras mutant mouse skin tumors. We will use a Systems Biology approach to visualize the architecture of Ras signaling in whole tissues in vivo and in epithelial cells derived from mutant mice, to examine the relationships between inflammation and susceptibility to development of Hras- or Kras-driven malignancies. This comprehensive systems-based approach will reveal conserved features of squamous carcinoma formation that will first help us to understand the genesis of these neoplasms, and begin to formulate strategies for prevention or treatment.